KR20140010468A - System for spatial extraction of audio signals - Google Patents

Abstract

The sound processing system receives an audio input signal comprising at least two different input channels of audio content. The sound processing system analyzes the audio input signal to separate the audible sound source included in the audio input signal into a sound source vector. The separation of the audible sound source into the sound source vector may be based on the perceptual position of each audible sound source within the listener perception sound stage. The sound source vector may represent a spatial slice that spans a listener perceptual sound stage that may be processed individually and independently by the sound processing system. After processing, the sound source vector can be selectively combined to form an audio output signal having an output channel used to drive each loudspeaker. Since the audible sound source is separate and independent, the audible sound source can be included in any one or more output channels.

Description

Spatial extraction system of audio signal {SYSTEM FOR SPATIAL EXTRACTION OF AUDIO SIGNALS}

This application claims the benefit of priority from US Provisional Application No. 61 / 248,770, filed October 5, 2009, which is incorporated herein by reference.

The present invention relates generally to an audio system, and more particularly to a system for spatially extracting content of an audio signal.

It is well known to generate an audible sound from an audio signal using an acoustic system. The audio signal may be a pre-recorded audio signal or a live audio signal. Upon receiving the audio signal, the acoustic system can generate an audible sound by processing the audio signal and providing the loudspeaker with the audio signal, typically in amplified form. An example of a live audio signal is a live stage performance by bands such as singers and orchestras. Examples of pre-recorded audio signals are compact discs or electronic data files in which songs of singers and bands are recorded. Other audio sources may also be similarly provided.

Typically, compact discs, electronic data files and other forms of audio signal storage consist of master recordings of audio sources such as singers and bands playing in studios or live concert scenes. Singers and bands can be played using microphones, amplifiers, and recording equipment to receive and capture the live music produced by the singers and bands. During recording, the sound mixing engineer can strategically position any number of microphones between the members of the band to receive the desired live sound for recording. The recording equipment includes any number of input channels configured to receive live audio inputs from microphones and other instruments played by the band.

The sound mixing engineer then mixes or adjusts the channel on which the audio signal was received to obtain the desired overall sound by the mantissa and the band. In addition, the sound mixing engineer can remix or otherwise adjust the recorded audio to specify how the recording will play back later. For example, the sound mixing engineer may have the violin perceived relative to the left side of the singer, and the guitar to the right side of the singer, such that when the recording is played through the loudspeaker of the audio system, the position of the singer recognized by the listener is at the center point. Individual audio signals can be adjusted to be perceived.

The audio system can also receive two or more channel audio input signals, such as stereo signals, and form more output channels than the received input channels. Such audio systems include those systems known as "Logic 7TM" and manufactured by Harman International Industries, a corporation of Northridge, California. Such systems distribute audio input signals to output channels based on paging of the audio input signals with respect to each other.

The sound processing system may receive an audio input signal comprising at least two separate audio channels. The audio input signal may be analyzed to determine the audible sound source or perceptual location of the audio source included in the audio input signal. The perceptual position can be identified based on the listener perception sound stage. The listener-aware sound stage is conceptually audio via a stereo audio system, or a surround voice audio system, or any other form of audio playback system capable of outputting audible sound to generate a listener-aware sound stage based on an audio input signal. It can be based on reproduction of the input signal.

The sound processing system may divide the listener perception sound stage into any predetermined number of perceptual positions (also termed spatial slices) of the listener perception sound stage. For example, if the audio input signal is a stereo input signal, the number of perceptual positions is the desired number of output audio channels, eg, left front output channel, right front output channel, center output channel, right side output channel, left side output channel, right side. There may be seven audio output channels representing the rear output channel and the left rear output channel. Also, the audio input signal can be divided into a plurality of predetermined frequency bands and the perceptual position of the audible sound source can be identified within the predetermined frequency band.

To separate the audio input signal into spatial slices, the sound processing system may determine and generate a gain vector for each spatial slice. Each gain vector includes a gain value that covers a predetermined frequency band within the entire frequency range of the audio input signal. The gain value may be generated based on the content of the audio input signal such that the audible sound source included in the audio input signal is separated into spatial slices according to the position of the audible sound source in the listener perception sound stage. The gain vector may be formed by a plurality of position filters forming a position filter bank. In one example, the number of position filters in the position filter bank may correspond to the number of spatial slices and the number of desired audio output channels.

The position filter bank may be applied to the audio input signal to divide the audio input signal into separate and independent sound source vectors such that each spatial slice may include a corresponding sound source vector. Each sound source vector may include an audio input signal portion representing one or more audible sound sources included in the spatial slice of the listener perceptual sound stage.

The sound source vector can be processed independently by the audio processing system. The processing may include a classification of an audible sound source included in each sound source vector. For example, the classification may include identification of an audible sound source that appears as a musical instrument, such as a trumpet, in the first sound source vector in the first spatial slice and identification of an audible sound source included in the second sound source vector in the second spatial slice as a human voice. can do. The processing may also include equalization, delay or any other sound processing technique.

After processing, the sound source vector can be combined to form an audio output signal that accommodates multiple audio output channels from which the loudspeakers can be driven. The combining may include combining the sound source vectors, dividing the sound source vectors, simply passing the sound source vectors as audio output channels, or generating an audio output signal to accommodate multiple audio output channels. Any other form of cooperative use.

Other systems, methods, features and advantages of the present invention will be or become apparent to those skilled in the art upon reviewing the following features and detailed description. All such additional systems, methods, features, and advantages are included within this description, are within the scope of the present invention, and are protected by the following claims.

The invention may be better understood with reference to the following figures and description. The components in the figures are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention. Moreover, in the drawings, like reference numerals designate corresponding parts throughout the different views.
1 is a block diagram of an exemplary audio system that includes an audio processing system.
2 is an example of a listener perception sound stage.
3 is another example of a listener perception sound stage.
4 is a graph illustrating an exemplary relationship between an estimated perceptual position and a listener perceptual sound stage.
5 is an example of a position filter bank.
6 is an example of a plurality of gain vectors in a listener perceptual sound stage and a plurality of spatial slices.
7 is an example of a block diagram of the audio processing system of FIG. 1.
8 is an example of another block diagram of the audio processing system of FIG.
9 is an example of another block diagram of the audio processing system of FIG.
10 is another example of a listener perception sound stage.
11 is an exemplary operational flow diagram of the audio processing system of FIG. 1.
12 is a second part of the operational flow diagram of FIG. 11.

1 is an example audio system 100 that includes an audio processing system 102. The audio system 100 may also include at least one audio content source 104, at least one amplifier 106, and a plurality of loudspeakers 108. Audio system 100 may be any system capable of producing audible audio content. Exemplary audio system 100 may be a fixed addressable consumer audio system, such as a car audio system, a home theater system, an audio system for a multimedia system, such as a theater or television, a public address system such as a multi-room audio system, a stadium, or a convention center. , An outdoor audio system, or any other site that wishes to reproduce an audible audio sound.

The audio content source 104 may be any form of one or more devices capable of generating and outputting different audio signals on at least two channels. Examples of audio content sources 104 include media players such as compact discs or video disc players, video systems, radios, cassette tape players, wireless or wired communication devices, navigation systems, personal computers, MP3 players or codecs such as IPOD ™ or at least Any other form of audio related device capable of outputting different audio signals on two channels.

In FIG. 1, audio content source 104 generates two or more audio signals in each audio input channel 110 from source material such as pre-recorded audible sound. The audio signal may be an audio input signal generated by the audio content source 104 and may be an analog signal based on analog source material or may be a digital signal based on digital source material. Thus, the audio content source 104 may include signal conversion capabilities, such as an analog to digital or digital to analog converter. In one example, audio content source 104 may generate a stereo audio signal consisting of two substantially different audio signals representing right and left channels provided in two audio input channels 110. In another example, audio content source 104 generates more than two audio signals on more than two audio input channels 110, such as 5.1 channels, 6.1 channels, 7.1 channels, or the like, or each of the same number of audio input channels ( Any other number of different audio signals generated at 110 may be generated.

Amplifier 106 may be any circuit or standalone device that receives an audio input signal of relatively small amplitude and outputs a similar audio signal of relatively large amplitude. Two or more input signals may be received on two or more amplifier input channels 112 and output on two or more audio output channels 114. In addition to amplifying the amplitude of the audio signal, the amplifier 106 also has a signal processing capability to change the phase, adjust the frequency equalization, adjust the delay, or perform any other form of manipulation or adjustment of the audio signal. It may include. The amplifier 106 may also include the ability to adjust the volume, balance and / or fade of the audio signal provided in the audio output channel 114. In an alternative, the amplifier may be omitted, for example, when the loudspeaker 108 is in the form of a set of headphones or the audio output channel functions as an input to another audio device. In another example, loudspeaker 108 may include an amplifier, for example when loudspeaker 108 is a standalone loudspeaker.

The loudspeaker 108 may be located in a listening room, such as a room, a vehicle, or any other room in which the loudspeaker 108 may operate. The loudspeakers 108 are of any size and can operate over any range of frequencies. Each audio output channel 114 may supply a signal to drive one or more loudspeakers 108. Each loudspeaker 108 may include a single transducer or multiple transducers. The loudspeakers 108 may also be operated in different frequency ranges such as subwoofers, woofers, mids and tweeters. Two or more loudspeakers 108 may be included in the audio system 100.

Audio processing system 102 may receive an audio input signal from audio content source 104 on audio input channel 110. After processing, the audio processing system 102 provides the processed audio signal on the amplifier input channel 112. The audio processing system 120 may be a separate unit or combined with the audio content source 104, the amplifier 106 and / or the loudspeaker 108. Also in another example, the audio processing system 102 includes an audio content source 104, an audio amplifier 106, a loudspeaker 108, and / or any other device or mechanism [other audio processing system 102]. Communication via a network or communication bus.

One or more audio processors 118 may be included in the audio processing system. The audio processor 118 may process audio and / or video signals, such as a computer processor, microprocessor, digital signal processor, or any other device, a series of devices or other mechanisms capable of performing logic tasks. It may be one or more computing devices. The audio processor 118 may be operated in conjunction with the memory 120 to perform instructions stored in the memory. This indication may be in the form of software, firmware, computer code, or some combination thereof and may provide the functionality of the audio processing system 102 when performed by the audio processor 118. Memory 120 may be any form of one or more data storage devices, such as volatile memory, nonvolatile memory, electronic memory, magnetic memory, optical memory, or any other type of data storage device. In addition to the instructions, operating parameters and data may also be stored in the memory 120. Audio processing system 120 may also include an electronic device, an electromechanical device, or a device for conversion between an analog and digital signals, a mechanical device such as a filter, a user interface, a communication port, and / or any other operational functionality. And a user and / or a programmer within the audio system 100.

In operation, audio processing system 102 receives and processes audio input signals. In general, during processing of the audio input signal, the audio processor 118 identifies each of the plurality of perceptual positions of the plurality of audible sound sources that appear within the audio input signal. The perceptual location represents the physical location of each audible sound source within the listener perception sound stage. Thus, if the listener is present in a live performance that occurs on the actual stage, the perceptual position is aligned with the position on the stage of the performer such as guitarist, drummer, singer and any other performer or object generating sound in the audio signal.

The audio processor 118 decomposes the audio input signal into a set of spatial audio streams or spatial slices, each containing audio content from each (at least) perceptual location. Any sound source co-located within a given perceptual location may be included in the same spatial audio stream. Any number of different spatial audio streams can be generated across the listener perception acoustic stage. The spatial audio stream can be processed independently by the audio processor 118.

In operation, the audio processor 118 may generate a plurality of filters for each of the plurality of output channels based on the identified perceptual positions of each audible sound source. The audio processor 118 may apply a filter to the audio input signal to generate a spatial audio stream. Spatial audio streams can be processed independently. After processing, the spatial audio stream can be assembled or otherwise recombined to generate an audio output signal having a plurality of respective audio output channels. An audio output channel is provided on the amplifier input line 112. The audio processing system 102 may provide more or less audio input channels than the number of input channels included in the audio input signal. Alternatively, audio processing system 102 may provide the same number of audio output channels as provided as input channels.

2 is an example illustrating recognition over a listener recognition acoustic stage 200 formed using a stereo system configuration for receiving audio input signals, such as stereo audio input signals. In FIG. 2, the left loudspeaker 202 and the right loudspeaker 204 are driven by respective left and right channels of the audio content source to produce sound received by the listener at the listening position 206. In another example, additional channels and respective loudspeakers, loudspeaker positions, and additional / different size listening positions may be illustrated.

In FIG. 2, the listening position 206 is disposed at a central point 208 substantially between the loudspeakers 202, 204 such that the distances to each loudspeaker 202, 204 are substantially the same. In this example, three factors may be combined to allow the listener to determine the perceptual position of any number of sound sources within the listener perception acoustic stage 200 based on the audible sound emitted from the loudspeakers 202, 204. Can be. The factors include the relative amplitude levels of the sound sources in the left and right channels, the relative delay (arrival time) of the sound sources in the left and right channels, and the relative phases of the sound sources in the left and right channels.

If the level of the sound source is perceived at the listening position 206 to be greater in the left channel (left loudspeaker 202), then the sound source is the first in the listener perceived sound stage 200 closer to the left loudspeaker 202. It is perceived by the listener to be placed in the perceptual position (S1) 210. Similarly, when the sound source first arrives at the listening position 206 from the right loudspeaker 204, the sound source is at a second perceptual position S2 (212) closer to the right loudspeaker 204. 200). Thus, depending on the intensity of the sound and the time of arrival, it may be perceived by the listener that different sound sources are at different perceptual positions in the listener perception sound stage 200. Also, if loudspeakers 202 and 204 are driven by an audio signal with a significant phase shift between them, it is perceived that the sound source is placed in a third perceptual position (S3) 214 past right loudspeaker 204. Can be. FIG. 2 is a simple illustration of some example points of a sound source within the listener perception sound stage 200, and in other examples, any number of sound sources may be provided disposed at any number of perceptual locations.

In FIG. 2, the listener aware acoustic stage 200 is divided into seven zones, also referred to as spatial slices or perceptual positions 218, 220, 222, 224, 226, 228, 230. In another example, the listener perception sound stage 200 may be divided into any other number of perceptual positions. In FIG. 2, the first perceptual position S1 210 is estimated by the audio processor 118 to be located in the third spatial slice 222, and the second perceptual position S2 212 is the fifth perceptual slice ( 226 is assumed to be located, and the central point 208 is assumed to be located at the fourth spatial slice 224.

3 is another example of a listener perceptual sound stage 300 decomposed into spatial slices. The acoustic stage 300 was formed using a surround sound system configuration to receive multi-channel audio signals, such as 5.1, 6.1, 7.1 or some other surround acoustic audio signal. In FIG. 3, the left speaker 302, the right speaker 304, the center speaker 316, the left speaker 308, the right speaker 310, the left rear speaker 312, and the right rear speaker 314 are listeners. It is located far from location 316. The listening position 316 is disposed at a substantially concentric point due to the circular position of the loudspeakers 302, 304, 306, 308, 310, 312, 314. In other examples, any other number of loudspeaker and / or loudspeaker positions as well as listener positions may be illustrated.

In FIG. 3, seven spatial slices or perceptual positions 320, 322, 324, 326, 328, 330, 332 corresponding to the respective loudspeakers 302, 304, 306, 308, 310, 312, 314 are shown. Surround listener position 316. In other examples, any number of spatial slices can be used. In addition, the width of each space slice may be different in various examples. For example, the spatial slices may overlap or be apart within the listener perception acoustic stage 300.

The audio processor 118 may estimate the first cognitive sound source S1 336 in the listener cognitive sound stage 300 to be disposed within the third spatial slice 324, and the second cognitive sound source S2 338. May be estimated to be disposed within the sixth spatial slice 330. In another example, any number of cognitive sound sources can be placed in spatial slices 320, 322, 324, 326, 328, 330, 332.

The estimation of the placement of the sound source within the listener-aware sound stage may be based on a comparison of the relative amplitude, phase and arrival time of the channels of the audio input signal. In the example of the stereo audio input signal consisting of the right channel R and the left channel L, the calculation of the point estimated by the audio processor is based on equation (1).

Where S (ω) is the estimated point in each listener perception acoustic stage 300, L (ω) is the complex representation (consisting of real and imaginary components) of the left audio input signal in the frequency domain and R (ω) ) Is a complex representation (consisting of real and imaginary components) of the right audio input signal in the frequency domain, and B is a balance function. V _L (ω) and V _R (ω) are separate complex vectors (consisting of real and imaginary components) each having the same magnitude as one. V _L (ω) and V _R (ω) can be used to apply a frequency dependent delay to L (ω) and R (ω). The value of the delay, and hence the values of V _L (ω) and V _R (ω), is offset so that any difference that may exist in the arrival time of a given sound source in the left (L) and right (R) input channels. Can be selected. Therefore, V _L (ω) and V _R (ω) can be used to time-align a given sound source in two input channels. It will be appreciated that the delay provided by V _L (ω) and V _R (ω) can alternatively be achieved in the time domain before converting the left and right audio input signals into the frequency domain. The variable ω represents the frequency or frequency range. The balance function may be used to identify whether the sound source in the listener-aware sound stage is to the left of the center of the listener-aware sound stage or to the right of the center of the listener-aware sound stage. The balance function B may be represented by Equation 2.

Here, A is an indication of amplitude comparison by the audio processor 118 of the magnitude of the left audio input signal L to the magnitude of the right audio input signal R. In one example, A may be set to 1 by the audio processor 118 when the amplitude of the left audio input signal is greater than the amplitude of the right audio input signal, and A is the amplitude of the right audio input signal. It may be set to 0 when it is equal to and may be set to −1 when the amplitude of the left audio input signal is smaller than the amplitude of the right audio input signal.

If there are multiple input channels, such as five or six input channel surround audio sources, other equations may be used instead of Equations 1 and 2 to account for multiple input channels.

Where S (ω) is the estimated point in each listener perception acoustic stage 300, M _k (ω) is a complex representation (consisting of real and imaginary components) of the k-th audio input signal in the frequency domain, V _k (ω) is a complex direction vector (consisting of real and imaginary components). C is an integer greater than 1 and represents the number of input channels, so in the case of five input channel surround audio sources, C = 5. The value of the direction vector V _k (ω) may be chosen to represent the angle of the speaker position as intended for the multichannel input signal. For example, in the case of a multi-channel input signal having five input channels, a typical playback configuration consisting of a center speaker disposed in front of 0 degrees, a left and right speaker at ± 30 degrees, and a left and right rear surround speaker at ± 110 degrees is employed. It is reasonable to assume that an input signal has been generated. In the case of the configuration of this example, a suitable choice of direction vectors is V _Center (ω) = 1 + 0i, V _Left (ω) = 0.866 + 0.5i, V _Right (ω) = 0.866-0.5i, V _LeftSurround (ω) = -0.342 + 0.940i, and V _{RightSurround} (ω) = -0.342-0.940i, where i is a complex operator that is the square root of -1. Equation 3 can be used to sum the contributions to the composite acoustic field from each of the input signal channels to derive a composite single vector. This composite signal vector is a complex value (consisting of real and imaginary components). The angle function in equation (3) can be used to calculate the angle of the composite signal vector resulting from the summing process. In calculating the angle in this example, the center channel speaker corresponds to 0 degrees. In another example, zero degrees may be placed elsewhere. The factor 2 / π approximates the value of S (ω) so as to be in the range between +2 and -2. Equation 3 can be used for an input signal having two or more channels.

Alternatively, in another example, multiple input channels can be split into pairs for application to Equations 1 and 2 such that multiple separate cognitive acoustic stages are created. For example, the cognitive sound stage may be created between a left front and a right front, between a left front and a left side, between a left side and a left rear, and the like. In another example, audio sources of three or more input channels may be downmixed to two input channel audio sources, such as downmixing five or six input channel surround audio sources to two input channel stereo audio sources. After extraction and processing, the audio source can be upmixed back to two or more audio output channels.

FIG. 4 shows the estimated position S (ω) calculated for the listener perception acoustic stage 404, eg, the listener perception acoustic stage 200 of FIG. 2; 402] is an exemplary graph illustrating the relationship therebetween. The listener aware acoustic stage may be divided into a plurality of predetermined zones, each having a position value from between the predetermined range of position values. In FIG. 4, the position value of the acoustic stage 404 is in a predetermined range of position values of −2 to +2, and the central position 406 at the central position of the listener perception acoustic stage 404 identified as the position zero zone. , Left side location 408 identified as -1 location zone, leftmost side location 410 identified as -2 location zone, right side location 412 identified as +1 location zone, and +2 location zone identified Included rightmost side position 414. In another example, another listener aware acoustic stage may be illustrated, such as the listener perceived acoustic stage shown in FIG. 3. In addition, other ranges of position values may be used to identify different zones across the listener perception acoustic stage, and there may be additional or fewer zones.

In FIG. 4, the estimated perceptual position S (ω); 402 is calculated to be between -2 and +2 to correspond to the points of the listener perceptual sound stage 404. In another example, estimated perceptual location [S (ω); Other values may be used to indicate 402. Estimated perceptual position [S (ω); Value is calculated according to Equation 1 to be +,-or 0 based on the amplitude comparison A (Equation 2).

Operation and signal processing in the audio system can occur in the frequency domain or in the time domain based on the analysis of the audio input signal. For brevity, this discussion focuses primarily on frequency domain based implementations, but time based implementations, or combination time based implementations and frequency based implementations are possible and within the scope of the system.

The audio input signal can be converted to a frequency domain representation by applying overlapping window analysis to a time sample block and transforming the sample using a Discrete Fourier Transform (DFT), waveform transform, or other transform process. Each block of time samples may be named a time instant or snapshot of the audio input signal. The time instant or snapshot can be any predetermined time period or time window. Thus, the audio input signal can be divided into snapshots or a sequence of consecutive or discontinuous segments, each segment having a start time and an end time constituting a predetermined amount of time between this start time and the end time. . The end time of one segment of the audio input signal may be adjacent to the start time of the subsequent segment of the audio input signal such that the segments are formed in an end-to-end configuration. In one example, each segment may represent a time window or snapshot with a duration of about 10 milliseconds. Typically, the snapshot has a duration of about 5 to 50 milliseconds. In the frequency domain, each snapshot of the audio input signal can be separated into a plurality of frequency bins across a predetermined frequency spectrum. The frequency bins may be of a predetermined size, such as about 50 Hz, respectively, to enclose a predetermined frequency range, such as an audible frequency range of 0 Hz to 24 kHz. For example, based on a predetermined sample rate, such as 48 kHz, and a predetermined number of bins, such as 1024 passes, each bin may have a bandwidth of 46.875 Hz. In another example, the size of the bin can be changed dynamically and automatically by the audio processing system based on the sample rate of the audio input signal. For example, if the audio input signal is a digital signal that can be sampled at a sample rate of 44.1 kHz, 48 kHz, 88.2 kHz or 96 kHz, then the sample rate of the audio input signal can be detected by the audio processing system and thus frequency The magnitude of may be adjusted such that the audio processing system is operated at the sample rate of the audio input signal.

In one embodiment, there may be 1024 frequency bins over an audible frequency range of 0 Hz to 24 kHz. Alternatively, a snapshot of the audio input signal can be partitioned into a frequency bank in the time domain using a bank of parallel band pass filters. The audio input signal may also be divided into a predetermined number of perceptual positions or spatial slices across the listener perception sound stage based on Equations 1 and 2. Within each of the perceptual positions, a segmented portion of the audio input signal can be displayed.

5 shows an exemplary position filter bank 500 generated by the audio processing system 102 based on the listener perception sound stage and equations (1) and (2). In FIG. 5, an indication of six position filters is illustrated. The position filter may match a number of output channels provided as audio output channels included in the audio output signal to drive the loudspeakers. Alternatively, any number of filters may be used to generate a corresponding number of output channels for further processing or use before forming an audio output channel to drive the loudspeakers. Thus, any number of position filters may be used and the output channels of the position filters may be further processed and then combined or split to match the number of audio output channels used to drive the loudspeakers. For example, if the audible sound source present in the audio output signal is not in the position of the listener aware acoustic stage corresponding to the audio output channel, two signals may be generated for the audio output channel on both sides of that position. In another example, if the audible sound source present in the audio input signal is in the position of a listener aware acoustic stage corresponding to two or more audio output channels, the signal may be doubled in the two or more audio output channels.

In FIG. 5, the position filter may include an output channel corresponding to the audio output signal. Thus, the position filter includes a center channel output filter 502, a right front output filter 504, a left front output filter 506, a right side output filter 508, a left side output filter 510, and a right rear output filter 512. ) And left rear output filter 514. In this example, the output filters 502, 504, 506, 508, 510, 512, 514 are the center, right front, left front, right side, left side, right rear and left rear designated loudspeakers, etc. in a surround sound audio system. It may correspond to an output channel driving each loudspeaker, and one or more speakers provide recognition above or below the listener's ear or any other speaker to provide the desired effect. In another example, output filters 502, 504, 506, 508, 510, 512, and 514 may correspond to intermediate output channels that are further processed to ultimately become part of two or more audio output channels. In other examples, fewer or more location filters may be provided and used as desired. The position filter bank 500 includes a first axis identified as the gain axis 518 and an estimated perceptual position S (ω); A second axis identified as the estimated perceptual location axis 520 corresponding to FIG. 4. In FIG. 5, the gain axis 518 is the vertical axis and the estimated perceptual position axis 520 is the horizontal axis.

Each filter may be configured and implemented by the audio processor 118 based on the estimated perceptual location of the sound source across the listener perceptual sound stage. The filter may be calculated by the audio processor 118 in the frequency domain or in the time domain based on the analysis of the audio input signal. In the frequency domain, using Equations 1 and 2, the estimated perceptual position can be calculated. As mentioned above, in one example, the calculated estimated perceptual position value may be a value between -2 and +2. In another example, any other range of values may be used for the calculated estimated perceptual position values. Based on the particular calculated estimated perceptual position, the corresponding gain value can be determined.

In FIG. 5, the intersection point 524 is at a gain value of about 0.5 on the gain axis 518. Intersection point 524 may mark the beginning of the transition of acoustic energy away from the first position and towards the second position. In the case of a position filter representing an output channel, the intersection point 524 may indicate a transition of acoustic energy between the first output channel and the second output channel. In other words, in this example, as the gain value in one channel decreases, the gain value in the other channel may increase accordingly. Thus, the sound output to the adjacently arranged output channels at any given point in time can be allocated between the adjacently arranged output channels based on the calculated estimated perceptual position values. For example, the center channel output filter 502 is at a gain of 1 when the calculated estimated perceptual position value is at 0, while the center channel output filter 502 is at a -0.5 calculated position. The gain value of is at about 0.15 and the gain value of the left front channel output filter 506 is at about 0.85. The intersection point 524 can be adjusted based on the filter structure characterized by the slope of the lines representing each position filter.

Thus, by calculating the estimated perceptual position 520 at the instant of time, the audio processing system can produce a corresponding gain value for the output filter during the same instant of time. As described above, the audio input signal is divided into frequency bands. Thus, the calculated gain value is calculated within each frequency band based on the application of Equations 1 and 2 to the portion of the audio input signal in each frequency band to calculate the estimated perceptual position 520. Intersection 524 shown in FIG. 5 may occur at gain values other than 0.5. The position filters 502, 504, 506, 508, 510, 512, 514 of the example shown in FIG. 5 overlap only with adjacent filters. Other position filter structures with more or less overlap between adjacent filters may be used. A position filter structure can be devised in which three or more position filters have a nonzero gain value for a predetermined estimated perceptual position S (ω) of the acoustic source across the listener perception acoustic stage. Additionally or alternatively, the gain value of the position filter can be positive and negative.

6 is an exemplary diagram of a listener perception acoustic stage 600 depicting a predetermined number x of perceptual positions or spatial slices 602 across the listener perception acoustic stage 600 at a timely time. As noted above, seven spatial slices are shown, but any number (x) of spatial slices 602 are possible. In FIG. 6, the listener perception acoustic stage 600 includes a left loudspeaker 604 and a right loudspeaker 606 that are generally symmetric about a center 608. In another example, other configurations of the listener perception acoustic stage can be implemented, such as the surround acoustic listener perception stage shown in FIG. 3.

As described above, Equations 1 and 2 apply to audio input signals divided into predetermined frequency bands or frequency channels. The gain value can also be derived as described above based on the calculated estimated perceptual position value. The gain value may be included in the position filter represented by the gain position vector 610 for each spatial slice of spatial slice 602. Each gain position vector 610 may include a gain value 612 such as a gain value ranging from 0 to 1.

In Figure 6, gain value 612 is represented by Gsn, where "s" is a spatial slice number and "n" is a frequency band location in each gain location vector 610 corresponding to a frequency bin number. Each gain position vector 610 may represent in a vertical direction the frequency range of the audio input signal from the first predetermined frequency f1 to the second predetermined frequency f2 within a specific spatial slice of the spatial slice 602. The number of gain values 612 in each gain position vector 610 may correspond to the number of the frequency bin Bn in which the audio input signal is divided. As discussed above, the audio input signal may be divided into a predetermined number of frequency bins, such as 1024 passes, over a predetermined range of frequencies f1 to f2, such as 0 Hz to 20 kHz. Thus, in one example, each gain position vector 610 may include 1024 gain values 612 (n = 0 to 1023) over a frequency range of 0 Hz to 24 kHz, which means that the audio input signal is 48 kHz. Can result in a gain value for each predetermined portion of the bandwidth (or frequency bin), such as about 46.875 Hz wide increments of the entire frequency range.

In operation, an audio input signal can be applied to the gain position filter. At each instant of time, each gain value 612 in each gain position vector 610 may be multiplied with a portion of the audio input signal In in the corresponding frequency bin Bn as follows.

Here, S _sn is an acoustic source value of the space slice number "s" corresponding to the frequency channel number "n".

The resulting acoustic source vector Ss formed from the array of acoustic source values Ssn in each spatial slice can populate each acoustic source in the spatial slice 602 for a moment in time. Each sound source value (“n” sound source value) of the sound source vector Ss may be distributed over a predetermined frequency range f1 to f2 according to the frequency bin Bn, similarly to the gain value. Thus, the frequency range of the sound source in a particular spatial slice 602 can be fully represented over the predetermined frequency range of f1 to f2. In addition, all sound sources present across the listener perceptual acoustic stage 600 in the audio input signal across the " s " spatial slice 602 in any given band of frequency corresponding to the frequency bin Bn. Can be represented. Since the gain value 612 is applied horizontally across the listener perception acoustic stage for the same frequency bin Bn, the gain value 612 is spatial slice (s) 602 in the predetermined frequency band n. If added over, the result may be a maximum gain value. For example, if the range of gain values is 0-1, the horizontal sum of gain values 612 across all spatial slices 602 for the first frequency bin B1 may be one.

The acoustic source vector Ss in each spatial slice 602 may represent one or more acoustic sources or audio acoustic sources across the listener perception acoustic stage. Audio input signals (audio source materials) could be generated or mixed by mixing engineers to place each sound source by perception. For example, the sound engineer may create (or play) a stereo audio recording such that when the audio recording is played through the audio system, the listener perceives that the listener is located near the front of the concert hall and a group of performers playing the instrument and singing near the center of the stage. Mix). In this example, the acoustic engineer is responsible for setting up a listener-aware sound stage such that the singer is located near the center of the sound stage, the bass guitar is located to the left of the listener-aware sound stage, and the piano is located to the right of the sound stage. Audio recordings can be mixed to distribute across. In another example, when an audio recording is created as a surround sound audio recording, the sound engineer is present in the audience and the listener is part of the audience in a concert hall where other listeners included in the recording are perceived to be behind and / or next to the listener. You may wish to be aware.

Each sound source can now be included in a separate sound source vector Ss in each spatial slice. Thus, adjustment and further processing of the individual sound sources can be performed by further processing the individual sound source vectors Ss. If the number of position filters in the position filter bank is the same as the number of audio output channels, each sound source vector Ss can be used as the source material to drive the loudspeakers. Alternatively, if the number of audio output channels is more or less than the number of sound source vectors Ss, the sound source vectors Ss are aggregated, combined, split, overlapped, passed, and / or comprise sound source vectors. It can be otherwise processed to generate an audio output signal to include each number of audio output channels. The audio output channel included in the audio output signal can also be further processed before being output to drive one or more respective loudspeakers.

7 is a block diagram example of a functional processing block of an audio processing system 102 operating in the frequency domain. The audio processing system 102 includes an audio input signal analysis module 700 and a post processing module 702. The audio input signal analysis module 700 includes an audio input preprocessing module 704, a sound source vector generation module 706, and a parameter input controller module 708. In another example, additional modules or fewer modules may be used to describe the functionality of the audio processing system 102. As used herein, the term “module” or “modules” is defined as software (computer code, instructions) or hardware (eg, circuits, electrical components and / or logic) or a combination of software and hardware.

In FIG. 7, the audio input preprocessing module 704 may receive an audio input signal. The audio input signal 712 may be a stereo pair of input signals, a multi channel audio input signal, such as a five channel, six channel or seven channel input signal, or any other number of audio input signals over two audio input signals. The audio input preprocessing module 704 may include any form of time domain for the frequency domain switching process. In FIG. 7, the audio input preprocessing module 706 includes a windowing module 714 and a converter 716 for each audio input signal 712. Windowing module 714 and converter 716 may perform window analysis overlapping on a block of time samples and converting the samples using a Discrete Fourier Transform (DFT) or other transformation process. In another example, the processing of the audio input signal may be performed in the time domain, and the audio input preprocessing module 704 may be omitted from the audio input signal processing module 700 and replaced by a time domain filter bank. .

The preprocessed (or not preprocessed) audio input signal may be provided to the sound source vector generation module 706. The sound source vector generation module 706 may generate the sound source generation vector Ss. The sound source vector generation module 706 may include a gain vector generation module 720, a signal classifier module 722, and a vector processing module 724. The gain vector generation module 720 may generate a gain position vector 610 for each of the spatial slices 602 as discussed with reference to FIG. 6.

The generation of the gain position vector by the gain vector generation module 720 may include the estimated position generation module 728, the position filter bank generation module 730, the balance module 732, the cognitive model 734, and the source model 736. And processing by the type detection module 738. The estimated location generating module 728 may calculate the estimated perceptual location value using Equation 1 as described above. The position filter bank generation module 730 may calculate the position filter bank 500 as described above with reference to FIG. 5, and the balance module may use Equation 2 to calculate the sound source generation vector Ss.

Cognitive model 734 and source model 736 may be used by the estimated position generation module 728, position filter bank generation module 730, and balance module 732 to improve processing to form gain position vectors. Can be. In general, the cognitive model 734 and the source model 736 are based on a snapshot-by-snapshot basis of the gain position vector to compensate for sudden changes in the calculated position of the audible acoustic source within the listener cognitive acoustic stage. Cooperative operation may be made to enable coordination in calculations. For example, the cognitive model 734 and the source model 736 can compensate for the presence of a particular sound source and the sudden change in amplitude in a listener cognitive sound stage that may otherwise cause a sudden change in the perceptual position. The cognitive model may perform smoothing of the gain position vector based on at least one of a time based auditory shielding estimate and a frequency based auditory shielding estimate during generation of the gain position vector over time (eg, over multiple snapshots). . Source model 736 can provide smoothing to monitor the audio input signal and avoid exceeding a predetermined rate of change in amplitude and frequency of the audio input signal over a predetermined number of snapshots.

Monitoring may be performed for each snapshot or for the timely instant of the audio input signal based on frequency bins considering at least one of the previous snapshots. In one example, two previous snapshots are individually weighted, averaged by a predetermined weighting factor and used for comparison to the current snapshot. The most recent previous snapshot may have a higher predetermined weight than the old snapshot. Upon confirmation by the source model 736 of the change in amplitude or frequency in excess of the predetermined rate of change, the cognitive model 734 is a source or audible sound included in the cognitive acoustic stage of the audio input signal, or the perceptual position of the audio source. The gain value of the gain position vector can be smoothed automatically and dynamically so as to reduce the rate of change of. For example, when multiple audio sources are sometimes at the same perceptual location or spatial slice and sometimes occupy different perceptual locations at different moments in time, smoothing may be used to avoid the audio source appearing to "jump" between perceptual locations. Can be. Otherwise such rapid movement between perceptual positions may be perceived by the listener as an audio source jumping from one of the loudspeakers driven by the first output channel to another of the loudspeakers driven by the second output channel. Can be.

Alternatively, or in addition, source model 736 may be a perceptual location or spatial slice in which perceptual locations may be automatically adjusted according to the audio source identified in the audio input signal based on the source included in source model 736. Can be used to define the boundary of the Thus, if the audio source is found to be at two or more perceptual positions, the area representing the perceptual position can be increased or decreased by adjusting the boundary of the perceptual position. For example, the region of the perceptual position may be widened by adjusting the intersection of the filters in the position filter bank 500 (FIG. 5) such that the entire audio source is in a single perceptual position. In another example, if two or more audio sources are determined to be at the same perceptual location, the perimeter location or boundary of the spatial slice may be gradually reduced until the audio source appears in a separate spatial slice. Multiple audio sources of a single perceptual location can be identified, for example, by identifying the source of the source model corresponding to the different operating frequency range of the identified source. The boundaries of other spatial slices can also be adjusted automatically. As discussed above, the boundaries of perceptual locations may overlap, distant, or be continuously aligned with each other.

The cognitive model 734 can also smooth over time the gain values contained in the gain position vectors to maintain a smooth transition from one moment in time to the next. Source model 736 can include models of different audio sources included in the audio input signal. In operation, the source model 736 can monitor the audio input signal and adjust the smoothing process by the cognitive model 734. As an example, source model 736 can detect a sudden onset of an acoustic source, such as a drum, and recognize model 734 to capture the onset of the drum at a unique location in space rather than dirty across the space slice. It is possible to reduce the amount of this smoothing. Using the model included in the source model 736, the cognitive model 734 can describe the physical characteristics of the sound source included in the audio input signal in determining how many predetermined frequency bands should be attenuated. Although shown in FIG. 7 as a separate model, cognitive model 734 and source model 736 may be combined in other examples.

The type detection model 738 can detect the type of audio input signal, such as classical music, jazz music, rock music, or story. The type detection module 738 may analyze the audio input signal to classify the audio input signal. Alternatively or additionally, type detection module 738 may be configured to determine externally provided information to determine and classify data included in the audio input signal, radio data system (RDS) data, or audio input signal as a particular type. Any other form can be received and decoded. The type information determined by the type detection module 738 may also be provided to other modules of the gain vector determination module 720. For example, in surround acoustic applications, the position filter bank generation module 730 receives an indication from the type detection module 738 that the type is classical music and automatically by adjusting the intersection of the filters in the position filter bank 500 (FIG. 5). Can be adjusted to prevent any part of the audio input signal from being output to the right rear and left rear audio output channels.

The signal classification module 722 may be operated at each perceptual location (spatial slice) across the listener perceptual sound stage to identify one or more audio sources included in each perceptual location of the perceptual locations. The signal classifier module 722 can identify the sound source from the sound source vector Ss. For example, at the first perceptual location, the signal classifier module 722 can identify each audio source as the singer's voice, and at the second perceptual location, each audio source is a particular instrument, such as a trumpet. In a third perceptual position, a number of individual audio sources can be identified, such as voice and a specific instrument, and in a fourth perceptual position of the listener perception acoustic stage, each audio source is audience noise such as clapping. Can be identified as Identification of the audio source may be based on signal analysis of the audible sound contained at a particular perceptual location.

The signal classifier module 722 determines the sound source based on the input information received from the parameter input controller 708, the output signal of the vector generation module 720, and / or the output signal of the vector processing module 724. You can do For example, the confirmation may be based on the frequency, amplitude and spectral characteristics of the sound source vector Ss in terms of position gain position vectors and parameters, such as the RDS data signal provided from the parameter input controller 708. Accordingly, the signal classifier module 722 may perform classification of one or more audio sources included in each of the perceptual positions in the listener perceptual sound stage. The classification may for example be based on a comparison with a library of predetermined sound sources, frequencies or timbre features. Alternatively or additionally, the classification may be based on frequency analysis, timbre features, or any other mechanism or technique for performing source classification. For example, the classification of the acoustic source may be based on the known distinctive features of the audio source, such as the relatively sudden drum onset characteristics, extraction and / or analysis of echo content included in the input signal, use of noise estimates included in the input signal, Detection of speech included in the input signal, detection of a particular audio source included in the input signal may be based.

The signal classification module 722 can cause the vector processing module 724 to assign a given sound source in a given spatial slice for a given output channel. For example, the speech signal may be assigned to a given output channel (eg, a central output channel) regardless of whether the speech signal is placed in a listener perception sound stage. In another example, a signal identified as a conversational pitch (eg, a story) may be assigned to two or more output channels, for example to increase clarity, or to obtain a desired acoustic field for any other reason.

In FIG. 7, the classification of the spatial slices is classified into feedback audio classification for each of 1) position filter bank generation module 730, 2) recognition module 734, 3) source model 736, and 4) type detection module 738. May be provided as a signal. The feedback audio source classification signal may include identification of each perceptual position across the listener perceptual acoustic stage and identification of one or more audio sources included in each perceptual position. Each module may use the feedback audio source classification signal when performing each processing of subsequent snapshots of the audio input signal.

For example, the position filter bank generation module 730 may locate and / or the width of the output filter in the position filter bank to capture all or almost all of the frequency components of a given sound source within a predetermined number of space slices, such as a single space slice. By adjusting, the region of the perceptual position can be adjusted. For example, the position and / or width of the spatial slice may be adjusted to adjust the intersection of the filters in the position filter bank 500 (FIG. 5) to track and capture the identified audio source in the audio input signal, such as the audio source identified as the voice signal. Can be adjusted by The cognitive model 734 can use the audio source classification signal to adjust the occlusion estimate based on the predetermined parameters. Exemplary predetermined parameters include whether the sound source has a strong harmonic structure and / or whether the sound source has a sharp onset. The source model 736 can use the feedback audio source classification signal to identify the audio source in the spatial slice of the listener perceptual sound stage. For example, where the feedback audio source classification signal indicates a voice audio source at some perceptual locations and a music audio source at other perceptual locations, the source model 736 may be configured to differ between the audio and music based models of the audio input signal. Applicable to the perceptual position.

Signal classifier module 722 may also provide an indication of the classification of the spatial slice at classifier output line 726. The classification data output on classifier output line 726 may be in any format compatible with the receiver of the classification data. The classification data may include the identification of the spatial slices and the indication of the sound source (s) included in each spatial slice. The receiver of the classification data may be a storage device, a computing device, or other internal or external device or module with a database or other data retention and organization mechanism. The classification data may be stored in association with other data such as audio data from which the classification data is generated. For example, the classification data may be stored in the header or side chain of the audio data. Offline or real-time processing of individual spatial slices or the overallness of the spatial slices in one or more snapshots may also be performed using classification data. Offline processing can be performed by devices and systems with computing power. Once stored in connection with audio data, such as in a header or side chain, the classification data can be used as part of the processing of the audio data by other devices and systems. Real-time processing by another computing device, audio related device, or audio related system may also use the classification data provided on output line 726 to process the corresponding audio data.

The type detection module 738 can use the audio source classification signal to identify the type of audio input signal. For example, if the audio source classification signal indicates only voice at different perceptual positions, the type may be identified by the type detection module 738 as a story.

The gain vector generation module 720 may generate a gain position vector on the gain vector output line 744 for receipt by the vector processing module 724. The vector processing module 724 can also receive the audio input signal 712 as an omnidirectional audio feed on the audio input signal omnidirectional feed line 746. In FIG. 7, the omni-directional transport audio signal is in the frequency domain, and in another example, the vector processing module 724 can be in the time domain or operate in a combination of the frequency domain and the time domain, and the audio input signal is a vector in the time domain. May be provided to a processing module 724.

The vector processing module 724 applies mathematics to apply the gain position vector to the audio input signal (a forward feed signal) in each frequency bin to generate an acoustic source vector Ss for each spatial slice across the listener perception acoustic stage. Equation 4 can be used. Individual and independent processing of the sound source vector Ss may also be performed within the vector processing module 724. For example, the individual sound source vectors Ss can be filtered or amplitude adjusted before being output by the vector processing module 724. In addition, effects can be added to some of the sound source vectors Ss, for example additional echoes can be added to the singer's voice. The individual sound source vectors Ss may also be independently delayed or altered, reconstructed, enhanced or corrected as part of the processing by the vector processing module 724. The sound source vector Ss may also be smoothed or otherwise processed separately before being output by the vector processing module 724. Also, the sound source vectors Ss may be aggregated, for example combined or divided, by the vector processing module 724 before being output. Thus, the original recording can be "adjusted" to improve the quality of reproduction based on the level of the individual spatial slice adjustments.

After processing by the vector processing module 724, the processed sound source vector Ss may be output as a sound source vector signal on the vector output line 748. Each sound source vector signal may represent one or more separate audio sources from within the audio power signal. The sound source vector signal may be provided as input signals to the signal classifier module 722 and the post processing module 702.

The parameter input controller 708 may optionally provide parameter inputs to the gain vector generation module 720, the signal classifier module 722, and the vector processing module 724. The parameter input may be any signal or indication available by the module to influence, change, and / or improve processing to generate a gain position vector and / or processed sound source vector Ss. . For example, in the case of a vehicle, the parameter inputs may include external signals such as engine noise, street noise, microphones and accelerometers placed inside and outside the vehicle, vehicle speed, climate control settings, convertible opening or closing, volume of the acoustic system, RDS Sources of data, audio input signals such as compact discs (CDs), digital video decoders (DVDs), AM / FM / satellite radios, cellular telephones, Bluetooth connections, MP3 players, Ipod®, or audio It can include any other source of input signal. Other parameter inputs may include an indication that the audio signal was lossy perceptual audio codec, the codec type used (MP3, etc.), and / or the input signal was compressed by the encoded bit rate. Similarly, in the case of a speech signal, the parameter input value may include an indication of the speech codec type employed, an encoded bit rate, and / or an indication of speech activity in the input signal. In another example, any other parameter useful for audio processing may be provided.

Within the gain vector generation module 720, the parameter input value may provide information to the type detection module 738 to detect the audio input signal. For example, if the parameter input value indicates that the audio input signal is input from the cellular phone, the type detection module 738 may indicate that the audio input signal is a voice signal. The parameter inputs provided to the signal classifier 722 can be used to classify individual audio sources in the spatial slice. For example, if the parameter input indicates that the audio source is a navigation system, then the signal classifier 722 looks at the spatial slice that contains speech as the audio source and ignores other spatial slices. In addition, the parameters may cause the signal classifier 722 to recognize noise or other audio content included in a particular spatial slice by the audio source. The vector processing module 724 can adjust the processing of the spatial slices based on the parameters. For example, in the case of a vehicle, a speed parameter can be used to increase the amplitude of a low frequency audio source, or a specific spatial slice, or a particular sound source vector at a higher speed.

In FIG. 7, the sound source vector signal may be processed through the post processing module 702 to switch from the frequency domain to the time domain using a process similar to the preprocessing module 704. Thus, the post processing module 702 may include a converter 752 and a windowing module 754 for the sound source vector signal. Converter 752 and windowing module 754 may use a Discrete Fourier Transform (DFT) or other conversion process to convert blocks of time samples. In another example, different frequency domains may be used for the time domain inversion process. In another example, the sound source vector signal provided to the vector output line 748 may be in the time domain due to processing by the sound source vector processing module 706 performed at least in part in the time domain. The sound source vector signal or post-processed sound source vector signal represents an audio source divided into spatial slices and can be subjected to further processing, can be used to drive the loudspeakers in the listening space, or any other audio processing related activity. Can be used for

8 is a block diagram of another example of an audio processing system 102 that may include an audio input signal analysis module 700, an acoustic source vector processing module 802, and a post processing module 804. The audio input analysis module 700 may include a preprocessing module 704, a sound source vector generation module 706, and a parameter input controller 708. The sound source vector generation module 706 may also include a gain vector generation module 720, a signal classifier module 722, and a vector processing module 724, as described above.

In FIG. 8, the preprocessing module 704 receives the audio input signal 806 in the form of a left stereo signal L and a right stereo signal R. In FIG. In another example, any number of audio input signals may be provided. The audio input signal 806 may be converted into the frequency domain by the preprocessing module 706 as described above, or may be directly received by the sound source vector generation module 706 in the time domain.

The sound source vector generation module 706 also uses the gain vector generation module 720, the signal classifier module 722, and the vector processing module 724, as described above, on the vector output line 748. (Ss) can be generated. The sound source vector Ss on the vector output line 748 may be received by the sound source bettor processing module 802. The sound source vector processing module 802 may also receive an audio classification signal from the signal classifier module 722 that directs the identification of the audio source in each spatial slice (sound source vector Ss).

The sound source vector processing module 802 may generate an audio output channel on the output channel line 810 based on the processed sound source vector Ss. The sound source vector processing module 802 may include a sound source vector modification module 812 and an assembly module 814.

The sound source vector modification module 812 may include functionality similar to that described above with respect to the vector processing module 724. The sound source vector modification module 812 includes a plurality of modification blocks 813 that can operate individually on each of the processed sound source vectors Ss. Thus, sound source vector correction module 812 adds echoes, performs equalization, adds delays, adds effects, performs dynamic range compression or expansion, improves transients, and extends signal bandwidth. Can be inserted and / or extrapolated to reconstruct the missing signal component, and / or perform any other audio processing related activities based on a per sound source vector Ss. Processing within the sound source vector modification module 812 may be used to correct, recover, and enhance the demoted audio signal. Thus, individual spatial slices across the listener perceptual sound stage can be independently modified, adjusted and / or compensated without affecting any other audio source in other sound source vectors Ss. For example, the delay of a particular spatial slice can be performed to emphasize the recognition of a particular spatial slice or to change the perceived width of the perceived acoustic stage.

The sound source vector modification module 812 may also perform modification of the individual sound source vector Ss based on the identification of the audio source in the individual vector. As noted above, the signal classification module 722 may operate at each perceptual location across the listener perceptual sound stage to identify one or more audio sources included in each perceptual location of the perceptual locations. After identification of the audio source, the corresponding sound source vector Ss can be modified based on the identified audio source. Unlike the vector processing module 724, which uses the identification of the audio source as feedback for processing after the snapshot, the acoustic source vector modification module 812 provides the identification of the audio source as omnidirectional transport. Thus, the sound source vector modification module 812 may process the individual sound source vector Ss based on the identification of each audio source as provided by the signal classification module 722.

Modifications based on the identification of the audio source may include correction of the individual audio source, adjustment of the perceived acoustic stage and / or the individual audio source included in the input signal, adjustment of the level of reflections, adjustment of the level of the speech source, Reduction or elimination of speech sources, enhancement of collision sources, dynamic range compression or expansion, bandwidth extension, extra and / or insertion to reconstruct lost components of individual audio sources, audio source specific effects or enhancements, and listener-aware sound stages Perception location adjustments may be included. Correction of the individually identified audio source may include replacement of the audio output portion of the particular audio source from a library or other audio source regeneration device, such as a MIDI player. For example, an audio source identified as a saxophone containing notes with noise output at a particular frequency may be replaced with the same note at the same frequency of saxophone audio output from the library or from a source capable of reproducing the saxophone's audio. The input audio signal may be corrupted or demoted as a result of processing by a cognitive audio codec, such as an MP3 codec, or any other form of lost compression. Other degradation / damage sources include insufficient audio recording and / or storage habits, AM / FM and satellite radio broadcasts, television broadcasts, wireless connections such as video codecs, Bluetooth, voice codecs, as well as telephone networks, including cellular networks.

The audio source specific effect or enhancement may comprise a change to the sound source value contained in the special sound source vector Ss specific to the identified audio source. For example, an audio source identified as speech can be increased in amplitude or adjusted to a specific frequency band to make it easier for the listener to recognize the speech. The particular sound source vector Ss can be compressed by the application of a dynamic range compressor to increase the intelligibility of the audio source appearing in the two or more sound source vectors Ss. For example, if the speaker voice is present in adjacent left and right sound source vectors that also include each instrument or background noise as well as the center sound source vector Ss, the center sound source vector is dynamically compressed or its level changed. Can be. In another example, an instrument, such as a trumpet, in a particular sound source vector Ss may be equalized to improve sharpness.

Perceptual position adjustment may include moving the identified audio source from one position to another in a sound field perceived by the listener. For example, a sound source, such as a singer's voice, may be located in the central channel with a second sound source, such as a guitar, present in the adjacently arranged sound source vector Ss at a voice stage perceived by the listener. Once identified by the signal classification module 722 as the singer's voice and other, the other sound source may be moved further away from the singer's voice by the sound source vector changing module 812 at the sound stage perceived by the listener. have. For example, the guitar may be moved to the right loudspeaker side by moving the audio source to another sound source vector Ss that has been identified as not containing an audio source by the sound source vector changing module 812. Vector processing module 724 is operative to identify and / or isolate acoustic sources and spatial slices as best as possible, while acoustic source vector change module 812 changes the identified and / or isolated acoustic sources and spatial slices. It is used to

The generation of the output channel may involve combining or separating the plurality of sound source vectors Ss from the assembling module 814 depending on the perceptual position from which the sound source vector Ss originates or the position in the user perceived acoustic stage of the spatial slice. It may include. For example, in a system with five output channels, the sound source vectors Ss from several perceptual locations adjacent to the center of the listener perception sound stage may be combined to form a central output channel that drives the center loudspeaker. In a five-channel surround sound output system according to another example where there are only four spatial slices, two of the spatial slices can be combined to form a lateral or rear output channel. In another example, the number of perceptual locations or spatial slices may match the number of output channels. As mentioned above, this allows two-channel stereo recording to be switched to 5, 6, 7, or any number of output channels.

The sound source vector Ss is rearranged or rearranged by an assembly module 814 that operates in conjunction with the sound source vector changing module 812 to move the audio source from the raw audio input signal to another location in the listener perception sound stage. Can be mapped. Since each audio source in the listener-aware sound stage may be included in a separate one of the sound source vectors Ss, the sound source may be moved or mapped to another location in the listener-aware sound stage. In other words, the sound source is determined and obtained in an audio input signal at the listener perception acoustic stage of each audio source, and the audio source is separated into individual perceptual positions or spatial slices by the sound source vector Ss. It can be determined whether is to be located at substantially the same location in the output audio channel or to be moved to a new perceptual location in the output audio channel.

For example, if the first perceptual location or spatial slice contains the mantissa's voice, and the second perceptual location located adjacent to the first perceptual location includes the other, the mantissa's voice is assigned or mapped to the central output channel. The guitar may also be assigned or mapped to both the left and right sides of the listener perception acoustic stage separated from the singer's voice. The singer's voice and the guitar properly map the sound source vector Ss containing the singer's voice to the central output channel and the sound source vector Ss including the guitar through the assembly module 814 on the left, right, front, side and / or the like. Or by mapping to a rear output channel. Thus, the audio processing system 102 can convert a two-channel audio input signal into any number of multichannel output signals, such as a surround sound output signal, as well as any of the output channels desired by an individual audio source in the audio input signal. May be assigned to one or more channels of.

In addition, the sound source vector Ss may be divided into two different sources such that, when driving the loudspeakers of which the output channel is adjacently arranged, the audio source included in the sound source vector Ss is perceptually perceived as being located between the two loudspeakers. Can be assigned to an output channel. Also, in special applications such as where loudspeakers are positioned at different heights and orientations within a vehicle, such as a door panel, dashboard, or tail deck of a vehicle, the acoustic source vector Ss is the listener's experience in the driver's and passenger seats of the vehicle. Can be selectively assigned proportionally with respect to the position of the loudspeakers to optimize In addition, the group of sound source vectors Ss may be fixedly mapped to one or more output channels. Alternatively, the sound source vector Ss may be an external parameter from the parameter input controller 708, the content of the audio input signal or the sound source after another sound source vector Ss appears in one or more output channels for a predetermined time. Operation may be grouped by the assembly module 814 to be automatically moved according to any other criteria useful for causing a change in the mapping of the vector Ss to the output channel. Thus, the mapping of the sound source vector Ss to the output channel may be a one-to-one, one-to-many, or many-to-one mapping. The mapping of some or all of the sound source vectors Ss is such that the left input signal is mapped to the left output channel of the playback speaker array (and subsequently to the speakers), and the right input signal is output channel to the right of the playback speaker array. Can be mapped to (and subsequently to speakers). Additionally or alternatively, the mapping of some or all of the sound source vectors Ss may be such that the left input signal is mapped to an output channel on the right side of the speaker array, and / or the right input signal is output to an output channel on the left side of the speaker array. Can be mapped. Additionally or alternatively, the mapping of some or all of the sound source vectors Ss may be such that the left input signal is mapped to output channels on both sides of the speaker array, and / or the right input signal is output to both output channels on the speaker array. Can be mapped. The choice of mapping may be predetermined and set by the user as needed to obtain the desired listener-acoustic sound stage for the output signal. The mapping of the sound source vector Ss to the output channel may be frequency dependent such that the mapping may vary with frequency. In one example, frequency dependent mapping can be used to obtain a better and more stable spatial image at the acoustic stage being reproduced.

The audio output channel on the output channel line 810 may be received by the post processing module 804. The post-processing module 804 can convert the frequency based audio output channel to the time output audio output channel using any form of frequency domain-time domain switching process. In FIG. 8, the post processing module 804 includes a converter 816 and a windowing module 818 for each of the audio output channels included in the audio output signal. Converter 816 and windowing module 818 may convert blocks of time samples using Discrete Fourier Transform (DFT) or other transform processing. In another example, the audio output channel provided to the output channel line may exist in the time domain due to processing by the sound source vector processing module 706 performed at least in part in the time domain, and the post processing module 804 may May be omitted.

9 is a block diagram of another example audio processing system 102 that may include an audio input signal analysis module 700 and a system management module 902. As described above, the audio input signal analysis module 700 may include a preprocessing block 704, a sound source vector generation module 706, and a parameter input controller 708. The sound source vector generation module 706 may also include a gain vector generation module 720, a signal classifier 722, and a vector processing module 724. Based on the audio input signal 904, the audio input signal analysis module 700 may generate a sound source vector Ss at the vector output line 748. In FIG. 9, the audio input signal 904 is illustrated as a left / right stereo pair provided in the time domain. In another example, there may be any number of audio input signals in the frequency domain or time domain.

The sound source vector Ss present in the vector output line 748 may be received by the system management module 902. The system management module 902 may include an energy measurement module 906 and a system control module 908. The energy measurement module 906 may include a vector measurement module 910 for receiving each sound source vector Ss at the vector output line 748. The vector measuring module 910 may measure energy levels of respective sound source vectors Ss. The vector measurement module 910 can measure signal levels using methods such as root-means-square (RM) based means or peak based means. Additionally or alternatively, the vector measurement module 910 can measure the loudness of the perceived signal.

System control module 908 can include controller 912, user interface 914, and data storage module 916. The controller 912 may be a standard processor similar to the processor 120 described in connection with FIG. 1 or may indicate functionality performed by the processor 120 (FIG. 1). User interface 914 can include any visual, audio and / or tactile mechanism process or mechanism that enables a user to provide and receive information from audio signal processing system 102. For example, the user interface 914 may include a display that converts electrical signals into information provided to the user in a predetermined visually recognizable form. Examples of displays include liquid crystal displays (LCDs), cathode ray tube (CRT) displays, electroluminescent displays (ELDs), head-up displays (HUDs), plasma display panels (PDPs), light emitting diode displays (LEDs), or vacuum fluorescent lights And a display VFD. The user interface 914 can receive and transmit, to the controller 912, an electrical signal indicative of the user's interaction with the audio signal processing system 102. In one example, the user interface 914 can include a user input device electrically connected to the controller 912. The input device may be a wheel button, joystick, keypad, touch screen configuration, or any other device or mechanism capable of receiving input from a user and providing such input to the controller 912 as an input signal. In another example, the display can be a touch screen display or any other module or device included in the audio signal processing system 102 that delivers signals to the controller 912. Information such as the area on the display the user touched, the length of time the user touched the display, the direction in which the user moves his or her finger relative to the display, and the like may be passed to the audio signal processing system 102 as another signal input.

The user interface 914 may include a voice-based interface that allows the user to acoustically interact with the audio signal processing system 102. The speech-based interface may allow a user to provide input to the audio signal processing system 102 using a microphone and speech recognition software. The user's voice may be converted into an electronic signal using a microphone and processed using speech recognition software to generate text data for the controller 912.

The data storage module 916 may include computer code that enables input / output logging and storage of data. The computer code may be in the form of logic and / or instructions executable by the controller 912. Execution of the command by the controller 912 may provide functionality to input and output energy levels from each of the respective sound source vectors Ss. In addition, the logging of any other data or parameters provided to or generated by the audio signal processing system 102 may be input and output by the data storage module 916. The data storage module 916 may include a database maintenance and control tool or any other form of data organization and storage device. The data storage module 916 may be included as part of the memory 118 described with reference to FIG. 1.

The audio processing system 102 of FIG. 9 may be used in conjunction with any other capabilities of the audio processing system 102 as described above. Thus, the user interface 914 can include functionality that allows a user of the audio processing system 102 to influence or control any functionality of the audio processing system 102 described above. For example, the user interface 914 allows the user to manually adjust the width and inclination of the individual position filters described with respect to FIG. 5. Thus, the user can manually adjust for which sound source vector Ss a particular sound source included in the audio input signal can be located, for example by a simple control knob. In another example, the user may have the ability to manually adjust how the sound source vector Ss is grouped, split or manipulated by the assembly module 814 described in connection with FIG. 8. Thus, if there are more sound source vectors Ss than the audio output channel, the user can adjust the perceived position of the audio source within the listener perception sound stage by adjusting the loudspeaker in which the sound source appears. In addition, manual input from the user to the parameter input controller module 708 described with reference to FIG. 7 may enter through the user interface 914. Manual adjustment of the audio source included in the sound source vector Ss or the sound source vector Ss using the sound source vector changing module 812 as described with reference to FIG. 8 may also be performed through the user interface 914. Can be. The output of the vector measurement module 910 of FIG. 9 is used by the change block 813 of FIG. 8 to indicate the level of the audio source included in the processed sound source vector Ss or the processed sound source vector Ss. Can be adjusted. For example, the modification block 813 may amplify the energy level of the acoustic source vector Ss used to generate the surround audio signal based on the relative energy level measured by the vector measurement module 910.

10 shows an example of certain adjustments within the audio processing system 102 that can be made to the sound source vector Ss to achieve a particular effect. The adjustment capability may be done manually by the user, automatically by the processor, or by any combination of manual and automatic control. 10 includes a listener-aware audio output sound stage 1002 formed of output audio channels that drive a right loudspeaker 1004 and a left loudspeaker 1006. The listener aware audio output sound stage 1002 includes a central position 1008. In another example, the listener perception acoustic stage 1002 may be a surround acoustic stage similar to FIG. 3.

Central space slice 1010 is illustrated in FIG. 10. In another example, any other output channel can be similarly adjusted. The center space slice may be adjusted in place within the listener aware audio output sound stage 1002 as indicated by arrow 1012 by the sound source vector change module 812 or the assembly module 814. In addition, the width or length of the listener perception acoustic stage 1002 connected to the center space slice 1010 may also be changed by changing the inclination of the center space filter 502 in the position filter bank 500 (FIG. 5). 1014). Adjusting the slope of any spatial filter will change the intersection with the spatially placed spatial filter. Thus, in the example of FIG. 10, the adjustment of the tilt of the center space filter 502, which narrows the center space slice 1010, causes the left and right speakers 1004, 1006) can be moved toward all or one. Conversely, the adjustment of the tilt of the center space filter 502 to widen the center space slice 1010 can move closer to the center 1008 within the listener perception acoustic stage 1002.

Additionally or alternatively, the amplitude or size of the audio output channel can be adjusted independent of the case of any other audio output channel. In FIG. 10, the adjustment of the amplitude of the center output channel is illustrated by arrow 1016. The adjustment of the amplitude may be performed in the sound source vector processing module 802 (FIG. 8) by adjusting the amplitude of the sound vector included in the sound source vector Ss identified to be in the central output channel 1010. .

A particular example of such amplitude adjustment is in the area of video broadcasting that includes audio. Since many video broadcasts, such as television programming, include an audio dialogue at a central location within the listener's acoustic stage of the audio input signal, the user can receive television without changing other audio sources included in the audio input signal. Have the ability to amplify the dialogue of programming. Thus, a user such as a user with a hearing aid having difficulty hearing a conversation due to the background noise included in the audio input signal may select a sound source vector Ss related to the sound source vector Ss associated with the center space slice, for example, by a predetermined amount of 6 dB. Positive amplification effectively amplifies the conversation while keeping the amplitude of the remaining sound source vectors Ss almost unchanged. Once the sound source vector Ss associated with the center space slice is amplified, the sound source vector Ss can be recombined to form one or more output channels, such as a pair of stereo output channels. Alternatively or additionally, spatial slices other than the central spatial slice found to contain negative may be amplified. In addition, amplification may optionally be applied based on the confirmation of the absence of speech in the spatial slice in which speech already exists. In another example, an advertisement received in television programming that includes compression that augments announcer speech may be attenuated in amplitude.

In one example, the adjustment of the area of the spatial slice 1010 by adjusting the position and / or width of the spatial slice 1010 and by adjusting the amplitude or size of the audio output channel is performed by the audio source in the listener perception acoustic stage as described above. It may be performed automatically by the audio processing system based on the confirmation. Additionally or alternatively, such adjustments can be made manually. For example, a user may have a first adjuster, such as a rotary knob or other form of user interface, which may cause the position of the spatial slice 1010 to move or move back and forth across the listener perception acoustic stage 1002. The user may include a second adjuster for adjusting the width of the spatial slice 1010 and a third adjuster for adjusting the loudness of audio content in the spatial slice 1010. Thus, the user adjusts the first adjuster to move the spatial slice 1010 to the periphery within the listener perception acoustic stage 1002 to position one or more sources of audible sound, such as guitar, to some extent within the listener perception acoustic stage 1002. You can do that. Once located, the user can adjust the second adjuster to adjust the width of the spatial slice 1010 to completely surround one or more sources of audible sound within the spatial slice 1010. In addition, when the user has finished adjusting the area of the spatial slice using the first and second adjusters as desired, the user adjusts the third adjuster so that the loudness of one or more sources of audible sound now enclosed within the space slice 1010. Can be increased or decreased.

11 is a flowchart showing an example of audio processing by the audio processing system 102 described with reference to FIGS. 1-10. In this example, the audio signal is provided in the time domain and converted into the frequency domain. In another example, the audio signal may be received in the frequency domain, and / or the processing may be done in the time domain and the frequency domain, only the time domain, or only the frequency domain. In block 1102, the audio processing system 102 receives an audio input signal from an audio source. The moment of time of the audio input signal is converted from the time domain to the frequency domain and divided into a frequency bin at block 1104. At block 1106, an estimated perceptual position S (ω) of each sound source through the listener perceptual sound stage may be determined by the estimated position generation module 728. The estimated position perception position S (ω) may be determined based on Equations 1 and 2 described above. The estimated perceptual position S (ω) is applied to the position filter bank 500 at block 1108.

At block 1110, a gain value is obtained for each frequency bin to form each position gain vector for one of a plurality of predetermined or user selected spatial slices. At block 1112, the cognitive model 734 and the source model 736 are applied to the gain position vector. At block 1114, it is determined whether the gain position vector has been formed for all spatial slices. If the gain position vector is not determined for all spatial slices, then the next spatial slice is selected at block 1116 and blocks 1110, 1112, 1114 are repeated. If the gain position vector at block 1114 has been determined for all spatial slices, the operation proceeds to block 1118 to form an acoustic source vector Ss for each spatial slice. The portion of the audio input signal in each frequency bin is multiplied by the corresponding gain value in each one of the gain position vectors to produce an acoustic source value Ssn to generate the acoustic source vector Ss for each spatial slice. Can be formed.

At block 1120, it is determined whether the acoustic source vector Ss has been determined for each spatial slice. Otherwise, at block 1122, the operation moves to the next spatial slice in which the acoustic source vector Ss has not yet been determined and blocks 1118 until the acoustic source vector Ss is derived for each spatial slice. 1120). If the acoustic source vector Ss has been derived for all spatial slices at block 1120, the operation proceeds to block 1124 of FIG. At block 1124, each acoustic source vector Ss is analyzed by the signal classifier 722 to identify the acoustic source, each represented by one or more acoustic source vectors Ss.

At block 1126, it is determined whether a sound source for each spatial slice has been determined. If not all spatial slices have been analyzed for the acoustic source, the operation returns to block 1124 for signal classification module 722 to identify additional acoustic sources in the spatial slice. On the other hand, once the spatial slices have all been considered, a feedback audio source classification signal is generated for each of the sympathetic slices, so that the location filter bank generation module 730 recognizes, at block 1128, to be used for the continuous snapshot processing of the audio input signal. May be provided to the model 734 and the source model 736.

In block 1130, a feedforward audio source classification signal is provided to the sound source vector change module 812 to further process the sound source vector Ss of the snapshot of the audio input signal currently being processed. The sound source vector changing module 812 may change the sound source vector Ss based on the forward transport audio source classification signal at block 1132. The sound source vector Ss may be reassembled with the assembly module 814 at block 1134 to form an audio output signal including the audio output channel. At block 1136, the audio output channel can be switched from the frequency domain to the time domain. The operation may then return to block 1104 to convert another snapshot of the audio input signal and perform the operations again.

Using the above-described audio processing system, by dividing an audio input signal into several spatial slices through a listener perceptual sound stage, any audio input signal of two or more channels is analyzed and the perceptual position of the audio source included in the audio input signal. You can check. The current snapshot of the audio input signal can be cut into several spatial slices, each containing an acoustic source vector (Ss) to identify the audio source. Once the audio sources are divided into sound source vectors Ss, each of the audio sources can be classified and further processed based on the classification. Alternatively, the audio input signal divided into spatial slices and each spatial slice including the sound source vector Ss may be processed independently. For other systems, this division is not possible to process the portion of the audio input signal representing the individual sound source. Once the independent processing of the individual spatial slices is performed, the spatial slices can be further manipulated to form the output audio channel. Manipulation may include moving, combining or dividing the spatial slices to form an audio output channel.

While various embodiments of the invention have been described, it will be apparent to those skilled in the art that many more embodiments and implementations are possible within the scope of the invention. Accordingly, the invention is not to be restricted except in light of the attached claims and their equivalents.

Claims (32)

As an audio processing system:
A processor;
Executed by the processor to analyze an audio input signal intended to drive a plurality of loudspeakers in a listening space, and to estimate a plurality of respective perceptual positions in a listener perception acoustic stage of each of the plurality of audio sources included in the audio input signal. Possible,
A gain vector generation module executable by the processor to generate a position filter bank comprising a plurality of position filters based on individual perceptual positions in the listener perception acoustic stage;
A vector processing module executable by the processor to apply the position filter bank to the audio input signal to produce a plurality of sound source vectors each representing one of each perceptual position
Audio processing system comprising a. 2. The sound source vector processing module of claim 1, further comprising a sound source vector processing module executable by the processor to modify the sound source vector and to combine the sound source vector to generate an audio output signal configured to drive a plurality of loudspeakers. And the sound source vector processing module is selectively coupled to the sound source vector such that the number of audio channels in an audio input signal is less than, more than, or equal to the number of channels in an audio output signal. Configurable audio processing system. 3. The position filter of claim 1 or 2, wherein each of the position filters comprises a plurality of gain position vectors, each of the gain position vectors having a plurality of gain values, each of the gain values being part of a total frequency range of the audio input signal. And the gain value is applied to the portion of the total frequency range of the audio input signal when the position filter bank is applied to the audio input signal. 3. The method of claim 1 or 2, wherein the analysis of the audio input signal by the gain vector generation module comprises a plurality of frequency passes each comprising the audio input signal and a band of frequencies included in the audio input signal. audio processing system comprising dividing into bins). 3. The audio processing of claim 1 or 2, wherein the gain vector generation module is further executable to use perceptual positions in the predetermined period to develop corresponding position gain vectors for each of the position filters in a predetermined period. system. 3. The audio processing system of claim 1 or 2, further comprising a signal classification module executable by the processor to identify each of the audio sources within each perceptual location. The method of claim 1, further comprising: generating a source model executable by the processor to smooth the audio input signal to prevent changes in amplitude and frequency of the audio input signal over a predetermined number of snapshots over a predetermined ratio. An audio processing system further comprising. The system of claim 1, wherein the position filter and the sound source vector are generated repeatedly at each of a plurality of time points, and the audio processing system is perceptual executable by the processor. Further comprising a model and a source model, the source model being executable to identify a change in amplitude or frequency of the audio input signal that exceeds a change in a predetermined rate, the perceptual model exceeding a change in the predetermined rate. And to dynamically smooth the gain position vector included in each of the position filters based on the identified change in amplitude or frequency. 8. A method according to any one of the preceding claims, characterized in that it identifies a sound source vector representing a predetermined one of the perceptual positions and adjusts the amplitude of the sound source indicated in the identified sound source vector. A sound source vector processing module executable by the processor to adjust the gain of the sound source vector, wherein the sound source vector processing module is further configured to generate an audio output signal to be provided to a plurality of loudspeakers; An audio processing system executable to combine the adjusted sound source vector with the rest of the sound source vector. As an audio signal processing method:
Receiving an audio input signal configured to drive a plurality of loudspeakers in the listening space using an audio processor;
Using the audio processor, identifying a plurality of perceptual positions of each of a plurality of sources of audible sound provided within the audio input signal, wherein the perceptual positions are physical locations of each source of audible sound in a listener perceptual sound stage. Indicating, the identifying step;
Using the audio processor, generating a plurality of filters for each of a plurality of individual output channels based on the identified perceptual positions for the respective source of audible sound;
Using the audio processor, applying the filter to the audio input signal to generate a plurality of sound source vectors each representing a portion of the audio input signal.
Wherein the audio signal processing method comprises the steps of: 11. The method of claim 10, further comprising modifying each of said sound source vectors separately and independently to modify said portion of said audio input signal separately and independently. 12. The audio signal of claim 11, further comprising processing the modified sound source vector using the audio processor to produce an audio output signal adapted to drive a separate loudspeaker in each of a plurality of separate audio output channels. Treatment method. 13. The method of claim 12, wherein processing the modified sound source vector comprises combining the sub-combinations of the modified sound source vectors together to form one respective audio output channel. The method according to claim 12 or 13, wherein the audio input signal comprises a plurality of audio input channels, and wherein the number of the individual audio output channels is larger or smaller than the number of the audio input channels. The method of any one of claims 10 to 13, wherein the step of identifying a plurality of perceptual positions divides the audio input signal into frequencies of a plurality of predetermined bands, and at least one of the frequencies of the predetermined band. Identifying a perceptual location of one or more sources of the plurality of sources of audible sound at a frequency of. 16. The method of claim 15, wherein identifying the perceptual positions of one or more of the plurality of sources of audible sound comprises each of a plurality of predetermined regions forming a listener perception sound stage from a predetermined position value of a predetermined range of position values. Audio signal processing method comprising: assigning to. 11. The method of claim 10, wherein applying the filter to the audio input signal comprises separating a group of sources of audible sound coupled to the audio input signal into a plurality of different sound source vectors. And the group of sources of is separated based on the identified individual perceptual position for each of the sources of audible sound in the group. A computer readable medium containing instructions executable by a processor:
Instructions for receiving an audio input signal configured to drive a plurality of loudspeakers in the listening space;
Instructions for generating a plurality of gain position vectors, each gain position vector corresponding to a position in the perceptual acoustic stage generated when the audio input signal is output as an audible sound in a listening space, and each gain position vector corresponding to the corresponding position of the gain position vector; Said vector generating instruction comprising a gain value at each of the frequencies of each of said plurality of predetermined bands of frequencies of said audio input signal at said position;
Generating a plurality of position filters for each of a plurality of individual output channels, wherein the position filters are generated from the gain position vectors;
Apply each said position filter to said audio input signal to form one acoustic source vector of a plurality of acoustic source vectors;
Computer readable medium comprising a. 19. The computer readable medium of claim 18, further comprising instructions for identifying an individual audio source in each said sound source vector, and instructions for separately and independently processing each said sound source in accordance with said identified individual audio source. Media available. 20. The method of claim 19, further comprising: combining the sound source vector to form an audio output signal comprising instructions for independently processing each said sound source vector and a plurality of audio output channels adapted to independently drive individual loudspeakers. Computer-readable media further comprising instructions. 21. The method of any of claims 18-20, wherein the instructions for generating a plurality of gain position vectors comprise instructions for converting the audio input signal into a frequency domain and dividing the audio input signal into frequencies of a predetermined band. Computer readable media. 21. The method according to any one of claims 18 to 20, wherein the command to apply each of the position filters to the audio input signal is such that each of the gain values at frequencies of one of the predetermined bands of frequencies is one of the predetermined bands. And instructions for applying to frequencies of a corresponding one of the frequencies of the predetermined band of an audio input signal. 21. The method of any of claims 18-20, wherein the instructions for generating a plurality of gain position vectors comprise instructions for generating a gain value for each of the frequencies of the predetermined band of an audio input signal at each of the positions. Computer-readable media comprising. As an audio signal processing method:
Receiving an audio input signal adapted to drive a plurality of loudspeakers in the listening space;
Dividing the audio input signal into a plurality of sound source position vectors using a position filter bank having a plurality of position filters constructed based on estimated perceptual positions of a plurality of sources of audible sound included in the audio input signal; Wherein each of said acoustic source position vectors represents a perceptual position over a listener perception acoustic stage, wherein at least some of said acoustic source position vectors comprise sources of audible sound included in said audio input signal;
Independently modifying the sound source position vector;
Combining the sound source position vectors to produce an audio output signal comprising a plurality of audio output channels each configured to drive a separate loudspeaker
Wherein the audio signal processing method comprises the steps of: 25. The method of claim 24, wherein dividing the audio input signal into a plurality of sound source position vectors further comprises dividing the audio input signal into frequencies of a plurality of predetermined bands, and generating a plurality of frequencies for each of the frequencies of the predetermined band. Generating a sound source value, wherein each of said sound source vectors is formed from said plurality of sound source values for a particular one of said perceptual positions. 25. The method of claim 24, wherein dividing the audio input signal into a plurality of sound source position vectors comprises applying the audio input signal to the position filter to generate the sound source position vector. . 25. The method of claim 24, wherein combining the sound source position vectors to produce an audio output signal comprises combining the sound source position vectors to form each of the audio output channels. 28. The method of any of claims 24 to 27, wherein combining the sound source position vector to produce an audio output signal comprises forming one of the audio output channels from one of the sound source position vectors. Audio signal processing method comprising a. 28. The method of any of claims 24 to 27, wherein combining the sound source position vector to generate an audio output signal comprises one of the sound source position vectors in at least two of the audio output channels. Audio signal processing method comprising the step. 28. The method of any one of claims 24 to 27, wherein independently modifying the sound source position vector comprises independently adjusting only the audio sound source included in a particular one of the sound source position vectors. Audio signal processing method comprising. 28. The method according to any one of claims 24 to 27, wherein the step of modifying the sound source position vector independently comprises: replacing the source of audible sound included in a first vector of the sound source position vector; Moving to two vectors. As an audio processing system:
Is configured to receive an audio input signal configured to drive a plurality of loudspeakers in the listening space;
Further configured to convert the audio input signal into a frequency domain;
Further configured to divide the audio input signal into a plurality of predetermined bands of frequencies;
In addition, it is configured to generate a location filter bank comprising a plurality of location filters each corresponding to one of a plurality of perceptual locations through a listener perceptual sound stage;
Further configured to apply the position filter bank to the audio input signal to divide a source of audible sound included in the audio input signal into a plurality of perceptual positions;
Further configured to separately process the source of the divided audible sound separately;
Further configured to combine the source of the divided audible sound to form an audio output signal comprising a plurality of audio output channels.
An audio processing system comprising a processor.

Images